ABSTRACT
Confidence intervals (CI) are based on standard errors. The 95" confidence interval, or 95" CI, is one of the most commonly reported confidence intervals. Sometimes researchers use other confidence intervals, such as the 68" or 99" confidence interval. The basic idea is the same for each kind of confidence interval. The only difference is that the larger the percentage used, the more confident one is that the true population value lies within the range. In addition, as the percentage goes up, the size of the interval typically increases. Because confidence intervals are estimates of the true values in populations, when the intervals for two groups on the same item overlap, researchers usually conclude that they failed to establish a reliable difference. Conversely, if the two intervals do not overlap, researchers feel more confident that there is a reliable difference between the groups. This makes comparing confidence intervals a type of informal significance test.
